Hi Rick,
We need to look at the current distribution and amplitudes on a 5/16
wave single wire folded counterpoise (FCP) quite carefully. It really
is NOT the same as two loaded 1/16 wave radials.
The deciding factor between the two is definitely not appearances.
When an FCP is either naturally in black or spayed with black, brown
and green against a background of trees or woods, our experience is
that it becomes invisible from the street, even when you know it's
there and where it is, actually up to HOA stealth levels. But
significantly more critical than that, if you really understood the
camouflage FCP's "Christine-approved" seal of approval both in nature
and depth of criticism, you would not propose appearance as a deciding
factor for two 1/16 radials. You'd be back to performance issues.
:>)
Just in general, if two loaded 1/16 wave radials were as good as the
experience we have had with the FCP, two loaded sixteenth waves would
have already been out there in number, tons of folks happy as a pig in
slop using it, it would be in ON4UN's book and we would not be having
this discussion. Even with the single inductor for both radials, a
pair of 1/16 radials requires something like a whopping 55
microhenries to tune it, and with a lossless coil has a 1.5:1 SWR
bandwidth of only 26 kHz. The coil required is almost all of a $70 ten
inch B&W #3026 Miniductor if you want to run QRO. Myself, the last
time I tried to do something that extreme I melted the plastic rods
with overheating.
The smallest successful radial set I have seen is ON4UN's proffer of 4
loaded 1/8 wave elevated radials, which is in use with good results at
one active contest station in my local area. John's radial length and
count choice, to my analysis anyway, seems a carefully chosen balance
between the loading inductor's size and losses, and narrowing of
bandwidth, and the loss multiplier of length over ground. But that
requires the 90 foot square for the radials, vs. the 66 foot line for
the FCP.
The FCP is not a shortened counterpoise like a pair of 1/16 radials.
The FCP is a LENGTHENED counterpoise, LONGER than than a quarter wave
in order to pull off a field cancellation trick by folding it. It
also has a useful side effect of moving a typical tuned resonant Z of
smallish vertical radiator plus counterpoise above 50 ohms.
The junction of the FCP and vertical radiator is driven with a 1:1
isolation transformer (required) between it and coax center conductor,
coax shield, and any connection to ground. A wind, rain and snow
static ground is provided by a 10 watt 5 meg resistor bridging the
windings which is the only connection between the two windings. The
length of the radiating wire is adjusted at the far end to present
zero reactance on the shack side of the isolation transformer. This
way the residual inductance of the isolation transformer and the
capacitive reactance of the FCP is soaked up in the pruning and
tuning. Back to the current distribution...
Remember that the current entering the FCP is set by the radiating
wire because the apparent series resistance in the FCP is so low
relative to the radiation resistance of the vertical radiator. The
FCP's beginning current would be the same amplitude as the beginning
current on the two radials. Set our imaginary power drive to get one
ampere at the base of the antenna in both cases.
Counting FCP segments 1 through 5. 33 feet per segment. Directions
used are for illustration only.
1: center to 33 feet east
2: 33 feet east back to center
3: center to 33 feet west
4: 33 feet west back to center
5: center to 33 feet east and end insulator.
Segments 2 through 5 carry the typical cosine current curve of a 0 to
90 degrees quarter wavelength. This is enforced working backward from
the end insulator. Segment 1 has the cosine of -22.5 degrees to 0
degrees. The current max is at the connection between segments 1 and
2.
Distributed current on segment 2 is the mirror image of distributed
current on segment 1. Segment two is reverse current direction from
segment one, therefore segments one and two completely cancel fields.
Distributed current on segment 3 minus distributed current on segment
4 (distributed effective current on the west side) is almost identical
to a mirror of distributed current on segment 5 which all that is left
uncancelled on the east side.
This means that the UN-cancelled distributed currents in the FCP are
pair of ramps either side of center corresponding to the cosine of
67.5 degrees through 90 degrees, PROPORTIONAL to the ramp currents on
two opposed elevated 1/16 wave radials, EXCEPT that on the two loaded
radials the ramp current begins at 1.0 relative, and on the FCP begins
at 0.38 relative at the center of the FCP.
Since the cancel at ground due to opposite currents near center is the
same in either system, and the ground loss is the integration of the
square of the field intensity at ground times the ground area, the
loss of the FCP is 0.145 relative to the radials' 1.00. So 1/16 wave
radials with perfectly loss-less loading coils inherently have seven
times more ground loss than the FCP fed equal current.
The actual dB dun to antenna system performance from a choice of
counterpoise depends on these effective series losses summed with the
feed Z of the vertical radiator. At my location up 90 and out 105
produces a tuned feed Z of 120 something. Choice of counterpoise will
not make nearly so much difference to me with that height and length
as to someone who can only get 50 feet vertical.
The rather large loading coil(s) needed to tune two 1/16 wave radials
on 160 puts one in a d**ned if you do and d**ned if you don't
situation. If the Q is high and loss low, the coils seriously reduce
the bandwidth of the antenna system, whereas the FCP leaves the
bandwidth quite more controlled by the vertical radiator. If the Q is
low and losses high, this adds coil wire loss to the already higher
ground loss of the two radials. Four times 1/8 wave elevated seems to
have this balanced.
Preliminary work seems to indicate that the FCP has better performance
than any elevated SPARSE radial system, except a virtual tie with
ON4UN's 4 times 1/8 wave elevated radial system with a lossless
inductor. BUT, and a big BUT, there is lot of work left to do to
PROVE that, best laid plans of mice and men, wishful thinking
syndrome, and Murphy all apply here. We will likely be able to do
some direct sky wave validation by balloon suspension of a string of
Elecraft XG3's driving impedance matched 1 meter vertical antennas on
frequencies separated by .5 kHz and run the tests in reverse
direction. Those 30's and 40's researchers would have killed to get
their hands on some of the stuff we have.
The real difference between those two is a 90 foot square and a 66
foot line for the property disadvantaged. If the comparison is
between ON4UN's 1/8 waves and a *PAIR* of FCP's at right angles fed in
parallel, there is no contest. 45 foot square beats 90 foot square
and should be visible on RBN if the radiators are low Z. (W0UCE pay
attention.)
Carefully controlled, the RBN comparisons seem to be good to within a
couple dB range. That's a lot less than the ten dB accuracy range
needed to positively identify really sucky radial systems as really
sucky.
73, and may your 160 antenna make you happy,
Guy.
On Wed, Nov 16, 2011 at 12:22 PM, Richard (Rick) Karlquist
<richard@karlquist.com> wrote:
> On 11/15/2011 9:10 PM, Guy Olinger K2AV wrote:
>
>> field, either buried or elevated. Therefore, presuming that undense
>> irregular radials that would fit would be excessively lossy, per RBN
>> data previously gathered, the +33, -33 foot linear folded counterpoise
>> (FCP) is used instead, elevated at 8 feet. The folds in the
>> counterpoise are designed to self-cancel fields as much as possible,
>> thereby minimizing ground induction, which is loss to skywave. The 66
>
>> 73, Guy.e
>
> It seems to me that the folded counterpoise is equivalent to
> a couple of loaded short radials, except that "linear loading"
> is used instead of lumping loading coils.
> Thus the ground induction loss is not reduced by the folding. So this is
> just a non-traditional implementation of 2 short loaded elevated
> radials. Nothing wrong with that, if implemented carefully. The decrease
> in gain is probably within the margin of error of RBN.
>
> In the described small backyard situation, I would think that making
> radials out of plain wire and loading them with coils at the feedpoint
> would be more acceptable from the visual clutter viewpoint.
>
> Rick N6RK
>
_______________________________________________
UR RST IS ... ... ..9 QSB QSB - hw? BK
|